Readers of Hilary Mantel’s Wolf Hall and its sequels know that however much the author has used her creative skills to fill in the many holes in our knowledge of Thomas Cromwell, she will have taken no liberties with the facts that are available to historians. I’m pretty confident that Ms Mantel could authoritatively lecture graduate students on Cromwell’s life and times, such is the depth of knowledge she acquired in the course of her research. To read these novels is to learn history – but in an infinitely more engaging manner than a dry textbook. Is this degree of accuracy and verisimilitude incumbent on all writers of historical fiction?

I would argue not.

One of my favourite novels is James Clavell’s magnificent Shōgun, set in feudal Japan and very loosely based on a real-life Englishman, Will Adams. However, Clavell made no attempt to accurately represent Adams, merely taking his story has a launching point for his imagined opus. Nonetheless, Shōgun is widely considered the greatest portrait of feudal Japan ever written – in fiction or non-fiction. The accuracy comes, not from fictionalising true events and people, but by maintaining integrity to the era and its customs and worldview.

In other words, Shōgun is not true, but it is accurate.

Wolf Hall and Shōgun are masterpieces, produced by writers who had steeped themselves in their respective periods. Sadly, we may not have the luxury of months and years to become equally knowledgeable about the place and time in which we are setting our novel. But we have to try. Diligent online research can turn up amazing resources on everything from food and drink in 18^th century New England to virtual reality tours of Augustan Rome or Renaissance Florence. I suspect that most writers of historical fiction are drawn to the genre by their love of research, so it is no sacrifice to spend days and weeks in study.

Nothing is more jarring in historical fiction than obvious anachronisms. However, it is important to be on guard for other, less obvious slip-ups. For example, no one would have answered ‘it’s half past two’ or ‘two-thirty’ in ancient Rome or in medieval England; no one thought it necessary to divide time so finely in the pre-industrial era. Console yourself with the thought that even Shakespeare got that one wrong: In Julius Caesar, Cassius announces: ‘The clock hath stricken three.’ There were no striking clocks in Julius Caesar’s Rome.

Even food and drink may have to be researched. A medieval serf wouldn’t have wolfed down his potatoes at lunchtime – potatoes weren’t introduced into England until around 1600. And unless he was by the bank of a river, he wouldn’t have washed his meal down with water – the principal safe drink available to him would have been weak beer. As our serf is by definition poor, he may well have had oysters for his meal; nutritious, plentiful and cheap. These are the details you can only pick up by thorough research.

Language itself can present the trickiest pitfalls. Unless you are prepared to undertake really serious research, slang is best avoided – even colloquial speech needs to be tackled with care. Just one example – ‘hi’ is first recorded as a greeting in American English in the 19^th century; but the Oxford Dictionary informs us that it is a Middle English expression in use 400 years earlier.

In order to avoid mistakes, some degree of immersion in your period is essential.  As always, The New York Review of Books and The Times Literary Supplement are excellent start points for identifying authoritative books on almost any topic in any period of history. The Library of Congress Catalog is staggeringly comprehensive and I think easier to use than the British Library catalogue. However, neither of these sources pass judgement on books, which of course is what we want; to identify the best guide to our topic rather than all available guides.

I am always very careful to avoid reading works of fiction set in the period I intend to write about. The danger is that you pick up a nugget of illuminating detail that you want to use, but in a work of fiction you can’t be sure if it is a historical fact or an imagined scene. Re-using a detail from a memoir or biography is quite different from reusing another writer’s creative invention.

The first is research, the latter is plagiarism.

For example, in my own book ‘Empress’, I repurpose an actual event from the life of Noël Coward, who appears as a character in several chapters. The story goes that an unsophisticated dinner guest at Coward’s table mistook a finger bowl for a drinking cup and drank from it. To avoid embarrassing his guest, and with a discreet glance at the others at the table encouraging them to follow his lead, Coward too drank from his bowl. I felt that this little incident gave a real insight to his personality. I reset the story on board a cruise liner and the unsophisticated man became an inexperienced young girl. Later on, I have a scene where Coward dedicates a composition to the young girl and gives her the manuscript containing his rough notes for the song. As this scene is entirely imagined, if another writer was to reuse or repurpose it, that would be simple intellectual theft.

My subsequent novel, ‘All Visible Things’, was set, in part, in the studio of Leonardo da Vinci. This presented various challenges. There is remarkably little authoritative information about Leonardo, the man. Was he gay? Was he a vegetarian? Did he have ADD? Books and PhD theses have been written on these questions. The good news from my point of view is that, in the presence of uncertainty, I was free to come down on either side of any argument.

However, there is broad agreement on where Leonardo was living at various points in his life. Therefore, in my story I did adhere to a strict timetable of the maestro’s sojourns from Florence to Milan to Roma and finally, to France.

In short, there is a genre, fictionalised history, which demands accuracy as far as possible; and more general historical fiction, where the rules are looser.